fraud-detection

Detect fraudulent orders using risk scoring, 3D Secure, velocity checks, and manual review queues.

3|1|Updated Mar 21, 2026
One-click install
npx skills add https://github.com/tomtoto757/ecomm-ai-team --skill fraud-detection-tomtoto757
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fraud-detection
Source: https://github.com/tomtoto757/ecomm-ai-team/tree/main/skills/platform-integrations-infrastructure/finsilabs/security-compliance/fraud-detection
Command: npx skills add https://github.com/tomtoto757/ecomm-ai-team --skill fraud-detection-tomtoto757

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires stripe, ioredis.

What problem does it solve? Payment fraud costs e-commerce merchants 2–3% of revenue through chargebacks, lost goods, and dispute fees. This Skill helps you layer fraud defenses — platform-native risk scoring, 3D Secure authentication, Redis-backed velocity checks, and manual review queues — so suspicious orders are blocked or held before fulfillment. ## Core Features & Use Cases - Platform-Specific Fraud Setup: Configure fraud tools for Shopify (Fraud Analysis, Signifyd), WooCommerce (Stripe Radar, Anti-Fraud plugin, Kount/NoFraud), BigCommerce, and custom headless storefronts. - Stripe Radar & 3D Secure Integration: Retrieve ML risk scores from PaymentIntents and request 3DS challenges for high-risk transactions to shift chargeback liability to the card issuer. - Velocity Checks & Review Queues: Implement Redis-based velocity limits per IP, email, and card fingerprint, plus an authorize-then-capture manual review workflow with Slack alerts and 48-hour auto-expiry. - Use Case: A store selling high-value electronics sees chargebacks climb past 0.5%. Use this Skill to enable Stripe Radar rules, add velocity checks that block cards used more than 3 times per hour, and route flagged orders into a manual review queue that holds payment capture until an analyst approves. ## Quick Start Use the fraud-detection skill to set up Stripe Radar rules and a Redis velocity check module for my WooCommerce checkout flow.

Frequently Asked Questions about fraud-detection

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up fraud detection for a WooCommerce store?▼

WooCommerce has no built-in fraud detection, so use Stripe Radar if you process payments through the WooCommerce Stripe gateway, or install the free WooCommerce Anti-Fraud plugin for rule-based risk scoring. High-volume stores can integrate Kount or NoFraud for enterprise fraud management.

How to implement velocity checks with Redis for checkout fraud prevention?▼

Use ioredis to increment counters per IP, email, and card fingerprint with EXPIRE set on first creation. Typical limits are 10 orders per IP per hour, 5 orders per email per 24 hours, and $500 daily spend per card fingerprint.

Does Shopify include built-in fraud analysis?▼

Yes, Shopify includes a free Fraud analysis indicator on every order showing risk level with specific reasons like card verification failure or IP/billing country mismatch. For advanced ML scoring and chargeback guarantees, install Signifyd from the Shopify App Store.

When should I use 3D Secure authentication on transactions?▼

Request 3DS challenges for high-risk transactions, such as when the Stripe risk score exceeds 70, because successful 3DS shifts chargeback liability to the card issuer. Use 'automatic' mode for low-risk orders to avoid checkout abandonment from unnecessary friction.

Why is my fraud system blocking legitimate customers?▼

A false positive rate above 1% means your rules are too aggressive. Whitelist B2B or high-LTV customers from velocity rules, use tiered limits based on account history, and track both fraud losses and revenue lost to false positives as KPIs.

What should happen to flagged orders in a manual review queue?▼

Flagged orders should be held with payment authorized but not captured, and the fraud team notified via Slack with the risk score. Auto-cancel unreviewed orders after 48 hours by cancelling the PaymentIntent, and never expose fraud rule details in client-facing error messages.